Mortality in Serologically Unconfirmed Mediterranean Spotted Fever [With Reply]

Home , Rickettsia rickettsii

University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

Other Publications in Zoonotics and Wildlife Disease Wildlife Disease and Zoonotics

2000

Mortality in Serologically Unconfirmed Mediterranean Spotted Fever [with Reply]

Pablo Yagupsky Ben-Gurion University of the Negev

Christopher D. Paddock Centers for Disease Control and Prevention, Atlanta, GA

James E. Childs Centers for Disease Control and Prevention, Atlanta, GA, [email protected]

Sherif R. Zaki Centers for Disease Control and Prevention, Atlanta, GA

Stephen A. Berger Tel Aviv Sourasky Medical Center, Tel Aviv, Israel

Follow this and additional works at: https://digitalcommons.unl.edu/zoonoticspub

Part of the Veterinary Infectious Diseases Commons

Yagupsky, Pablo; Paddock, Christopher D.; Childs, James E.; Zaki, Sherif R.; and Berger, Stephen A., "Mortality in Serologically Unconfirmed Mediterranean Spotted Fever [with Reply]" (2000). Other Publications in Zoonotics and Wildlife Disease. 48. https://digitalcommons.unl.edu/zoonoticspub/48

This Article is brought to you for free and open access by the Wildlife Disease and Zoonotics at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Other Publications in Zoonotics and Wildlife Disease by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. JID 2000;181 (February) Correspondence 809

Karin Galil,1 Rosalyn Singleton,3 and Grin Levine2 iterranean spotted fever (MSF), however, is milder than that 1National Immunization Program and 2Respiratory Diseases Branch, of RMSF, especially in children, and the fatality rate is con- National Centers Centers Disease Control for Infectious Disease, for sidered to be ^0Zo [2]. In 1993, Wolach and I [3] reported fatal and Atlanta, ^Arctic Prevention, Georgia; Investigations Program, MSF in 3 Israeli children who had with a febrile National Centers for Infectious Disease, Centers for Disease Control presented and Prevention, Anchorage, Alaska disease lasting 5-7 days, septic shock, mental changes, hypon- atremia, and bleeding tendency. A rash, macular or purpuric, was present in 2 children and absent in the third. A presumptive References diagnosis of meningococcemia or sepsis of unknown origin was 1. Dargan JM, Coplan PM, Kaplan KM, Nikas A. Reemergence of invasive entertained. The 3 children were empirically treated with 0- Haemophilusinfluenzae type b disease in Alaska: is it because of vaccination lactam antibiotics and aggressive supportive therapy but died with polyribosylribitol phosphate outer membrane protein complex (PRP- within 24 h of admission. The diagnosis of MSF was confirmed OMPC) or failure to vaccinate with PRP-OMPC? [letter]. J Infect Dis in 2 patients by cell culture and by animal inoculation in the 2000;181:806-8. third. Antibodies to as deter- 2. Takala AK, Santosham M, Almeido-Hill J, et al. Vaccination with Haemo- spotted fever-group rickettsiae, philus influenzae type b meningococcal protein conjugate vaccine reduces mined by microimmunofluorescence, were negative in 2 chil- oropharyngeal carriage of Haemophilus influenzae type b among American dren, whereas a borderline titer of 1 : 80 was found in 1. Indian children. Pediatr Infect Dis J 1993; 12:593-9. Early administration of specific therapy with tetracyclines or 3. Adegbola RA, Mulholland EK, Secka O, Jaffar S, Greenwood BM. Vacci- chloramphenicol significantly decreases mortality due to rick- nation with a Haemophilus influenzae type b conjugate vaccine reduces ettsial infections Institution of such re- oropharyngeal carriage of H. influenzae type b among Gambian children. [4]. therapy, however, J Infect Dis 1998;177:1758-61. quires that the rickettsial etiology of the illness be suspected 4. Smith-Vaughan HC, Priprakash KS, Leach AJ, Mathews JD, Kemp DJ. Low on clinical and/or epidemiologic grounds, because these anti- genetic diversity of Haemophilus influenzae type b compared to nonencap- microbial drugs are not usually given empirically to febrile pa- sulated H. in a in which H. is en- influenzae population influenzae highly tients. In addition, pediatricians frequently hesitate to start demic. Infect Immun 1998;66:3403-9. with in febrile children without a 5. Barbour ML, Booy R, Crook DW, et al. Haemophilusinfluenzae type b carriage therapy tetracycline young and immunity four years after receiving the Haemophilus influenzaeoligo- definitive proof of a rickettsial diagnosis, because of the po- saccharide-CRM197(HbOC) conjugate vaccine. Pediatr Infect Dis J 1993; tential risk of teeth staining. Usually the diagnosis of rickett- 12:478-84. sioses relies on serologic tests, whereas culture of the organism 6. Takala AK, Eskola J, Leinonen M, et al. Reduction of oropharyngeal carriage and polymerase chain reaction methods are not routinely used of Haemophilus influenzaetype b (Hib) in children immunized with an Hib or available. The series of Pad- conjugate vaccine. J Infect Dis 1991; 164:982-6. universally patients reported by 7. Mohle-Boetani JC, Ajello G, Breneman E, et al. Carriage of Haemophilus dock et al. [1] and our 3 patients clearly show that serologic influenzae type b in children after widespread vaccination with conjugate tests may be unreliable for diagnosing or excluding rickettsial Haemophilus influenzae type b vaccines. Pediatr Infect Dis J 1993; 12: infections, especially in persons presenting with fulminant dis- 589-93. ease. Had specific isolation techniques for rickettsiae not been 8. Jonsdottir KE, Steingrimsson O, Olafsson O. Immunisation of infants in Ice- the would have been missed in our 3 land against Haemophilus influenzae type b. Lancet 1992;340:252-3. attempted, diagnosis patients, and their deaths would have been attributed to infections The opinions expressed in this paper are those of the authors and do not caused by unidentified pathogens. Thus, it appears that, when reflect the views of the Indian Health Service. necessarily alternative methods for the disease are not used, Reprints or correspondence: Dr. Karin Galil, National Immunization Program, confirming Centers for Disease Control and Prevention, 1600 Clifton Rd., E-61, Atlanta, the diagnosis of MSF may be missed by serology, resulting in GA 30333 ([email protected]). underestimation of the true case/fatality rate of the infection. The Journal of Infectious Diseases 2000; 181:808-9 Because MSF may follow an unpredictable, rapid fatal course ? 2000 by the Infectious Diseases Society of America. All rights reserved. similar to that of it seems to advise 0022-1899/2000/18102-0071502.00 RMSF, prudent prompt administration of empirical tetracycline therapy whenever the diagnosis is suspected.

Pablo Yagupsky Mortality in Serologically Unconfirmed Clinical Microbiology Laboratory, Soroka Medical Center, Mediterranean Fever Spotted Ben-Gurion University of the Negev, Beer-Sheva, Israel

To the Editor?I read with interest the article by Paddock et al. [1], who described the detection of occult mortality due to Rocky Mountain spotted fever (RMSF) by demonstrating References Rickettsia rickettsii antigens or DNA in blood and tissues [1]. 1. Paddock CD, Greer PW, Farebee TL, et al. Hidden mortality attributable to In Mediterranean countries, Israel, fever is including spotted Rocky Mountain spotted fever: immunohistochemical detection of fatal, caused by members of the R. conorii complex, which are an- serologically unconfirmed disease. J Infect Dis 1999;179:1469-76. tigenically related to R. rickettsii. The clinical course of Med- 2. Raoult D, Weiller PJ, Chagnon A, Claudet H, Gallais H, Casnova P. Medi- 810 Correspondence JID 2000;181 (February)

terranean spotted fever: clinical, laboratory and epidemiological features and around blood vessels and within reticuloendothelial cells of 199 cases. Am J Med Trop Hyg 1986;35:845-50. (figure 1). 3. P, Wolach B. Fatal Israeli fever in children. Clin Infect Dis Yagupsky spotted We with recommendation to administer 1993;17:850-3. agree Yagupsky's 4. Woodward TE. Rocky Mountain spotted fever: epidemiological and early tetracycline antimicrobials (preferably doxycycline) to children clinical signs are keys to treatment and reduced mortality. J Infect Dis of any age when spotted fever-group rickettsioses are consid- 1984;150:465-8. ered in the differential diagnosis. The recognized propensity of tetracyclines to bind to dental enamel should not dissuade Reprints or correspondence: Dr. Pablo Yagupsky, Clinical Microbiology Lab- phy- oratory, Soroka Medical Center, Ben-Gurion University of the Negev, Beer-Sheva sicians from using the most effective drug for the treatment of 84101, Israel ([email protected]). these potentially life-threatening infections. The decision to use The Journal of Infectious Diseases 2000; 181:809-10 doxycycline in young children is never made casually, and the ? 2000 the Infectious Diseases of America. All reserved. by Society rights rationale for this choice of therapy should be discussed with 0022-1899/2000/18102-0072502.00 the child's parents [5]. However, it should be recognized that a single short course (i.e., 5-7 days) of doxycycline should not result in of teeth For Reply cosmetically significant staining [6, 7]. patients with Rocky Mountain spotted fever or severe MSF, there is a relatively narrow window of time during which ef- To the Editor?We appreciate the comments by Yagupsky [1], fective antibiotic therapy dramatically reduces the risk of death which underscore several of the salient features of spotted fe- [8]. Since laboratory tests available to most physicians do not ver-group rickettsial infections that we emphasized in our re- assist in early diagnosis, initiation of therapy should be based port [2]. For patients, diagnosticians, and clinicians faced with on clinical, and especially epidemiologic, findings (e.g., unex- the potentially tragic consequences of Rocky Mountain spotted plained febrile illness associated with known tick bite or tick fever or severe Mediterranean spotted fever (MSF), these points exposure or unexplained fever with thrombocytopenia, rash, or merit repeating. First, the nonspecific signs and symptoms early headache occurring during spring or summer months in an area in the course of rickettsial infections mimic many other infec- where the disease is endemic). tious and noninfectious syndromes. Second, the diagnostic chal- Isolation of R. conorii from patients' blood and tissues con- lenges posed by these diseases are compounded by the lack of firmed the diagnosis of spotted fever for each of the 3 pediatric rapid and widely available confirmatory tests early in the course patients described by Yagupsky and Wolach [3] and reflects the of the illnesses. Finally, spotted fevers have the potential to kill clinical acumen of these investigators. Although isolation is the otherwise healthy persons in ^ week after onset of symptoms, reference standard for diagnosis, culture is seldom attempted emphasizing the need for early administration of appropriate during the acute phase of illness, even when a rickettsial in- therapy. fection is suspected, and cannot be performed retrospectively The immunohistochemical (IHC) test described in our report on autopsy specimens unless samples were appropriately col- uses an antibody that reacts with multiple spotted fever-group lected and stored (e.g., frozen at -70oC). Biosafety level 3 prac- rickettsiae, including members of the Rickettsia conorii complex tices and facilities are recommended if culture of spotted fe- that causes MSF. In this context, we have used this assay to ver-group rickettsiae is attempted [9]. In contrast, IHC testing diagnose previously unexplained fatal illnesses caused by spot- provides a method for retrospective confirmatory diagnosis on ted fever-group rickettsiae other than R. rickettsii. Our expe- archived samples months or even years after the illness. In this rience reinforces the comments of Yagupsky [1] and includes context, IHC is a versatile technique that can help resolve prob- confirmation of fatal spotted fever infections in 2 adult Israeli lematic issues relating to the diagnosis of otherwise unexplained patients for whom serologic evidence of MSF was lacking. The illnesses and provide new insights into the clinical course and first patient was a 31-year-old woman who died 6 days after epidemiologic features of these diseases [10]. onset of an illness characterized by fever, myalgias, headache, and respiratory insufficiency. The second patient was a 38-year- Christopher D. Paddock,1 James E. Childs,1 old man who died 7 days after being hospitalized for fever and Sherif R. Zaki,1 and Stephen A. Berger2 vomiting. Both patients developed thrombocytopenia and pe- 1Centers for Disease Control and Prevention, Atlanta, Georgia; techial or rashes over the course of their illnesses. purpuric 2Tel Aviv Sourasky Medical Center, Tel Aviv, Israel Similar to 2 of the pediatric patients described in the earlier report by Yagupsky and Wolach [3], neither of these 2 patients References demonstrated diagnostic levels of antibody reactive with spotted fever-group rickettsiae when serum samples were tested by 1. Yagupsky P. Mortality in serologically unconfirmed Mediterranean spotted use of an indirect immunofluorescence assay. However, tissues fever [letter]. J Infect Dis 2000;181:809-10. 2. Paddock Greer Ferebee et al. Hidden attributable from both patients obtained at autopsy demonstrated micro- CD, PW, TL, mortality to Rocky Mountain spotted fever: immunohistochemical detection of fa- scopic lesions consistent with histopathologic of fatal findings tal, serologically unconfirmed disease. J Infect Dis 1999;179:1469-76. MSF [4], and IHC staining for spotted fever-group rickettsiae 3. Yagupsky P, Wolach B. Fatal Israeli spotted fever in children. Clin Infect revealed abundant rickettsial antigens and intact rickettsiae in Dis 1993;17:850-3. JID 2000;181 (February) Correspondence 811

Figure 1. Immunohistochemical localization of spotted fever-group rickettsial antigens in tissues of patients with fatal spotted fever, by immunoalkaline phosphatase stain with naphthol phosphate-fast red substrate and hematoxylin counterstain. A, Rickettsiae and rickettsial antigens associated with damaged endothelium in myocardial interstitium (patient 1); original magnification, X158. B, Abundant rickettsial antigens within denuded intravascular endothelial cells in small vessel in kidney (patient 1); original magnification, XlOO. C, Rickettsial nodule in cerebral cortex (patient 1); original magnification, X158. D, Spotted fever-group rickettsiae and rickettsial antigens in cytoplasms of Kupffer cells in liver (patient 2); original magnification, X158.

4. WalkerDH, Herrero-HerreroJI, Ruiz-BeltranR, Bullon-SopelanaA, Ra- children.Ann Pharmacother1997; 31:492^1. mos-HildagoA. The pathologyof fatal Mediterraneanspotted fever. Am 8. KirklandKB, WilkinsonWE, SextonDJ. Therapeuticdelay and mortality J Clin Pathol1987;87:669-72. in cases of Rocky Mountainspotted fever. Clin Infect Dis 1995;20: 5. AbramsonJS, Givner LB. Shouldtetracycline be contraindicatedfor therapy 1118-21. of presumedRocky Mountain spotted fever in childrenless than 9 years 9. RichmondJY, McKinneyRW, eds. Rickettsialagents. In: Biosafetyin mi- of age?Pediatrics 1990; 86:123^1. crobiologicaland biomedicallaboratories. 4th ed. Washington,DC: US 6. GrossmanER, WalchekA, FreedmanH, FlanaganC. Tetracyclinesand GovernmentPrinting Office, 19^9:148-52. permanentteeth: the relationbetween dose and tooth color. Pediatrics 10. ZakiSR, PaddockCD. Theemerging role of pathologyin infectiousdiseases. 1971;47:567-70. In:Scheld WM, Craig WA, Hughes JM, eds. Emerging infections 3. Wash- 7. Cale DF, McCarthyMW. Treatment of Rocky Mountainspotted fever in ington,DC: AmericanSociety for MicrobiologyPress, 1999:181-200. 812 Correspondence JID 2000;181 (February)

Reprints or correspondence: Dr. Christopher D. Paddock, Viral and Rickettsial HIV-1 load than by self-reported adherence, had a direct effect Zoonoses Centers for Disease Control and 1600 Clifton Branch, Prevention, Rd., on Candida colonization. It would be MS G-13, Atlanta, GA 30333. reducing oropharyngeal of interest to know if the association of Candida The Journal of Infectious Diseases 2000; 181:810-2 independent ? 2000 by the Infectious Diseases Society of America. All rights reserved. colonization with HIV-1 load was seen in the subgroup of pa- 0022-1899/2000/18102-0073502.00 tients known not to have received either current or recent protease inhibitor therapy. Until longitudinal studies are available and actual rather than reported use of chemotherapeutic agents with antifungal activ- Oropharyngeal Candida Colonization and Human ity can be accurately measured and controlled for in analyses, Immunodeficiency Virus Type 1 Infection it may be premature to conclude that HIV-1 load has a direct effect on local mucosal immunity, which promotes oropharyngeal Candida colonization. To the Editor?Gottfredsson et al. [1], in a cross-sectional study of oropharyngeal Candida colonization in 83 patients with hu- Robert S. Klein,1 Julia H. Arnsten,1 and Jack D. Sobel2 man virus 1 infection, immunodeficiency type (HIV-1) reported 1Departments of Medicine and of Epidemiology and Social Medicine, that only plasma HIV-1 load was predictive of Candida load; Montefiore Medical Center and Albert Einstein College of Medicine, neither CD4 lymphocyte level nor use of antiretroviraltherapy Bronx, New York; 2Department of Medicine, Wayne State University was associated with colonization [1]. They hypothesized that School of Medicine, Detroit, Michigan high levels of HIV-1 may suppress local mucosal immune re- sponses, thereby permitting increased Candida colonization, References and speculated that control of HIV-1 replication may restore local immune function. Their findings certainly are consistent 1. Gottfredsson M, Cox GM, Indridason OS, de Almeida GMD, Heald AE, Perfect JR. Association of plasma levels of human immunodeficiency virus with these possibilities. However, there are at least 2 alternative type 1 RNA and oropharyngeal Candida colonization. J Infect Dis 1999; explanations that might account for the interesting findings. 180:534-7. Cross-sectional studies do not provide information about 2. Gruber A, Lukasser-Vogl E, Borg-von Zepelin M, Dierich MP, Wiirzner R. causality for any associations found. However, even if high Human immunodeficiency virus type 1 gpl60 and gp41 binding to Candida levels of HIV-1 do cause increased Candida colonization, sup- albicans selectively enhances candidal virulence in vitro. J Infect Dis 1998;177:1057-63. pressed local mucosal immunity is not the only possible mech- 3. Haynes RB, McKibbon KA, Kanani R. Systematic review of randomized anism that such an occurrence. It was might explain recently trials of interventions to assist patients to follow prescriptions for medi- shown in vitro that HIV-1 glycoproteins may promote the vir- cations. Lancet 1996;348:383-6. ulence of Candida albicans [2]. Thus, increased local HIV-1 4. Eisen SA, Miller DK, Woodward RS, Spitznagel E, Prybeck TR. The effect replication might have a direct effect on increased Candidacolo- of prescribed daily dose frequency on patient medication compliance. Arch Intern Med 1990; 150:1881^1. nization, without necessarily requiring the presence of defects 5. Arnsten JH, Grant RW, Demas PA, Gourevitch MN, Schoenbaum EE. Ad- in local immune function. such a mechanism need Furthermore, herence with antiretroviraltherapy in HIV-infected drug users. J Gen Intern not imply disassociation of HIV-1 effects on local mucosal im- Med 1999; 14(Suppl 2):7. munity from those on CD4 lymphocyte counts, as would be 6. Cassone A, De Bernardis F, Torosantucci A, Tacconelli E, Tumbarello M, suggested by a better correlation of Candida colonization with Cauda R. In vitro and in vivo anticandidal activity of human immuno- virus J Infect Dis 180:448-53. HIV-1 effects on local immune function than with effects on deficiency protease inhibitors. 1999;

CD4 lymphocyte counts. Reprints or correspondence: Dr. Robert S. Klein, AIDS Research/Division of In addition, patient self-report is known to overestimate ad- Infectious Diseases, Montefiore Medical Center, 111 E. 210th St., Bronx, NY 10476-2490 ([email protected]). herence to prescribed medications [3], especially for regimens The Journal of Infectious Diseases 181:812 requiring multiple doses daily [4]. Preliminary observations of 2000; ? 2000 by the Infectious Diseases Society of America. All rights reserved. HIV-infected current and former drug users in the Bronx, New 0022-1899/2000/18102-0074502.00 York, showed that medication adherence assessed by electronic monitoring was significantly correlated with HIV load but that self-reported adherence was not [5]. HIV-1 load is inversely Reply correlated with actual use of effective highly active protease inhibitor-containing antiretroviral regimens. Even though oro- To the Editor?We thank Klein et al. [1] for their cogent re- pharyngeal Candida colonization was not independently as- sponse to our work [2] in which we reported that, among var- sociated with the reported use of protease inhibitors, its asso- ious clinical parameters in a cohort of human immunodefi- ciation with HIV-1 load suggests that there may well have been ciency virus (HlV)-infected patients, plasma HIV RNA levels less use of protease inhibitors in colonized patients. Since pro- had the most significant association with oropharyngeal Can- tease inhibitors have anticandidal activity [6], it is possible that dida colonization. As Klein et al. point out, this association the actual use of protease inhibitors, better reflected by low could be the result of 3 possibilities: HIV affecting mucosal